Abstract:The traditional on-line monitoring system for tower capsizing often has some shortcomings, such as data acquisition, single analysis, difficulty for non-professionals to understand and understand, and backward operation and maintenance management mode. In order to solve these problems, this paper synthesizes the related information of tower inclination, basic level, wind speed, wind direction and power by studying artificial intelligence technology. The machine learning algorithm is used to construct the tower overturning fault model, calculate the eigenvalues of the model reflecting the state of the equipment, detect the changing trend of the eigenvalues of the model, and realize on-line detection of the tower overturning state and early warning of the deterioration process. At the same time, this paper also focuses on the AI-based wind turbine tower overturning intelligent early warning system construction objectives, system architecture, diagnosis process and system development platform. Through this system, the ability of data mining technology is continuously improved, and the health status of the wind turbine is more accurate judgement and timely detection of hidden troubles, Safe operation of fan foundation and tower barrel is ensured, and major safety accidents such as inclination or collapse of fan are prevented.